Pathfinder
Editor, Senior Moderator
- Article
- Open access
- Published: 14 March 2025
Scientific Reports volume 15, Article number: 8797 (2025)
Cite this article
Abstract
The current outbreak of highly pathogenic avian influenza (HPAI) viruses of the H5N1 subtype clade 2.3.4.4b in dairy cattle in the United States has affected nearly 900 dairy farms and resulted in at least 39 human infections, putting health authorities and the scientific community on high alert. Here we characterize the virus growth properties and host-pathogen interactions of an isolate obtained from a sick dairy cow in Texas in vitro and in vivo and compare it to an older HPAI isolate. Despite so far being associated with mild disease in human patients, the cattle H5N1 virus showed superior growth capability and rapid replication kinetics in a panel of human lung cell lines in vitro. In vivo, cattle H5N1 exhibited more intense pathogenicity in mice, with rapid lung pathology and high virus titers in the brain, accompanied by high mortality after challenge via different inoculation routes. Additionally, the cattle H5N1 demonstrated efficient antagonism of overexpressed RIG-I- and MDA5-mediated innate antiviral signaling pathways. In summary, this study demonstrates the profound pathogenicity and suggests a potential innate immune escape mechanism of the H5N1 virus isolated from a dairy cow in Texas.
Introduction
Since 1997, highly pathogenic avian influenza (HPAI) A viruses of the A/goose/Guangdong/1/1996 (H5N1) lineage have often been the cause of devastating outbreaks in domestic and wild bird flocks, occasionally crossing the species barrier to infect mammals, including humans, and raising a concern that such viruses may eventually become more pathogenic and/or transmissible in humans[SUP]1[/SUP]. This concern has grown over the last few years, with frequent reports of outbreaks of clade 2.3.4.4b viruses in domestic and wild birds, infections in mammals in several countries and continents, and even more alarmingly, infections of large groups of wild and farmed mammals such as seals and minks, with evidence of possible mammal-to-mammal transmission[SUP]2,3,4[/SUP].
A whole new chapter in the history of HPAI started in March 2024, when an outbreak in dairy cows in the state of Texas was first reported. Affected cows presented with decreased feed intake, decreased rumination time, mild respiratory signs, lethargy, signs of dehydration, and abrupt drop in milk production with altered milk characteristics (i.e., a thickened, yellow appearance). Lethality is low, and most cows recover in a few weeks, although milk production may remain low for longer periods. Strikingly, infection is characterized by high tropism of the HPAI virus for the mammary gland tissue, resulting in a viral mastitis and substantial viral shedding in the affected milk[SUP]5,6[/SUP]. This current outbreak is highly unusual, as until now, cattle were generally seen as only infrequently infected by influenza viruses, most commonly influenza D. Though it has long been known that the bovine mammary glands are susceptible to experimental infection by influenza viruses upon intramammary inoculation[SUP]7[/SUP], spontaneous mammary infections with any influenza viruses had never been reported. Importantly, many farms reported concurrent deaths of domestic and peridomestic animals, most notably birds (grackles, pigeons) and outdoor domestic cats which are believed to have acquired infection likely through the consumption of contaminated raw milk from the sick cows[SUP]8[/SUP].
Although the first documented cases were reported in March, epidemiological information and phylogenetic analyses indicate the dairy cattle HPAI outbreak may have started several months earlier, in late 2023[SUP]9[/SUP]. Since then, it has quickly spread through the United States and, as of December 16, 2024, has reached 16 states, being confirmed in 853 dairy herds[SUP]10[/SUP], making it by far the most extensive and prolonged instance of mammal-to-mammal transmission of HPAI viruses known to date. As of December 26, 2024, CDC has confirmed 65 human H5 influenza infections in the United States, of which 39 were associated with infected dairy cows and 23 with exposure to infected poultry[SUP]11[/SUP], most of which have been benign. The continuous circulation of HPAI A(H5N1) in cattle is a major concern. It creates an ideal condition for adaptation of the virus to further increase the transmissibility among the mammalian hosts. Moreover, the proximity of cows and people also favors reassortment with human influenza viruses. Adaptation and reassortment are the two major mechanisms for the generation of pandemic influenza[SUP]12[/SUP].
To better understand this threat, an understanding of the growth properties and the host-pathogens interactions of the cattle HPAI A(H5N1) is paramount. In this study, we characterized the growth properties of HPAI A(H5N1) isolate A/cattle/Texas/56283/2024(H5N1) (cattle-HPAI) in vitro, and its pathogenicity, tissue tropism and infection kinetics in vivo, while comparing it with an older HPAI isolate, A/whooper swan/Mongolia/244/2005(H5N1) (M244/05)—a classical avian isolate that is known to be pathogenic in mammalian animals including mice, ferrets and non-human primates[SUP]13,14,15,16[/SUP]. Finally, we identified a potential mechanism by which cattle H5N1 isolate antagonizes the host innate immune response.
...
Discussion
The current global multispecies panzootic caused by HPAI clade 2.3.4.4b H5N1 viruses, which has spread to dairy cattle causing a nationwide outbreak in the United States, has the scientific community and public health authorities on high alert. Accordingly, many measures have been instituted to monitor, contain and stop HPAI A(H5N1) transmission, with seemingly little effect[SUP]4[/SUP]. These interventions have included increased virus surveillance in dairy herds, animal workers, and dairy products, tracking and limiting movement of lactating cows, stepping up biosecurity measures in dairy operations, and recently, mandatory pooled milk screening[SUP]32[/SUP].
Here we report results of our ongoing work for characterizing the host-pathogen interactions of the cattle HPAI A(H5N1) with mammalian hosts in vitro and in vivo.
Other groups have studied dairy cattle H5N1isolates in mouse models and compared its pathogenicity with other 2.3.4.4b clade viruses, as well as other more distant HPAI strains (e.g. A/Vietnam/1203/2004) and even human H1N1 viruses[SUP]17,18[/SUP]. Here we expand this comparative characterization by pairing the cattle-HPAI with yet another classical H5N1 isolate, M244/05. We chose this classical and well-characterized yet phylogenetically distinct strain of HPAI considering its relevance in the context of viral evolution, zoonotic potential and public health concern. Choosing a pure-avian strain to compare with cattle-HPAI provides an opportunity to study the basis of viral adaptation to cattle and other mammalian species, which could provide new insights into the risk of human infection and potential spread.
We show that the cattle HPAI A(H5N1) is highly lethal to mice, with an MLD[SUB]50[/SUB] of only 2.5 TCID[SUB]50[/SUB] upon intranasal challenge of BALB/c mice, and that even very low infectious doses can kill mice in only 4 days. Cattle-HPAI virus had faster replication kinetics especially in the brains of challenged mice, compared with M244/05—after challenging mice with 10 TCID[SUB]50[/SUB] of cattle-HPAI or 750 TCID[SUB]50[/SUB] of M244/05, the growth of the former overtook that of the latter by 2 DPI in the brains and 4 DPI in the lungs. In our hands, M244/05 showed limited replication in the brain, with virus being detected at low levels in only one of 3 animals at 4DPI, progressing to 2 out of 3 mice at 6 and 8 DPI, with maximum titers of 10[SUP]2[/SUP] and 10[SUP]3[/SUP] TCID[SUB]50[/SUB]/g respectively. This finding is consistent with a previous report[SUP]14[/SUP]. However, another report saw more robust virus growth of M244/05 in the brain of mice, with virus being detected in all animals and reaching mean titers of 10[SUP]4[/SUP] egg-infective doses (EID) at 9 DPI[SUP]13[/SUP].
We also show that the high pathogenicity of cattle-HPAI in mice is not strain- or age-specific, as challenge of older C57BL/6 mice led to similar outcome.
Upon histopathological analysis, lungs of cattle-HPAI-infected mice showed severe congestion and edema accompanied by vascular leakage, in the absence of inflammatory infiltration at 4 DPI, whereas lungs of M244/05-infected mice showed marked cell death and perivascular inflammatory infiltration, despite cattle-HPAI showing more extensive virus antigen staining at this time post-infection. This can be interpreted as a sign of early, overwhelming immunosuppression caused by the cattle isolate. Mice challenged with cattle-HPAI also showed intense congestion in the brain, which was not observed with M244/05, which is in keeping with the detection of virus antigens by IHC and the virus titers data in that organ.
Eisfeld et al. characterized the pathogenicity of a New Mexico isolate of HPAI A(H5N1), A/dairy cattle/New Mexico/A240920343-93/2024, in vivo using the same mouse model we report here. That isolate showed an MLD[SUB]50[/SUB] of 31.6 PFU —a lethal dose over one log higher than the Texas isolate used in the current study. Moreover, the Texas isolate characterized here showed much faster disease progression, with most mortality occurring at 4 DPI even at the lowest infectious doses, and with no mouse surviving past day 4 after challenge with 10[SUP]3[/SUP] TCID[SUB]50[/SUB] of virus, while similar challenge with the New Mexico strain would take twice as long (eight days) to kill all mice. In another study by the same group, it was found that two other strains of cattle H5N1, one isolated from a cow and another from a human patient, both in Texas, were both highly lethal at very low challenge doses such as 10 and even 1 PFU[SUP]33[/SUP], which is consistent with our findings.
The Texas isolate used in our study also showed faster replication kinetics in the brain: after challenge with only 10 TCID[SUB]50[/SUB], virus was detected in the brains of 2 out of 3 mice already at 2 DPI, with titers to the order of 10[SUP]3[/SUP] TCID[SUB]50[/SUB]/g of tissue, and 3 out of 3 mice at 4 DPI, all of them with titers above 10[SUP]7[/SUP] TCID[SUB]50[/SUB]/g of brain tissue. By contrast, as reported by Eisfield et al., 2024, after intranasal challenge of mice with 100 PFU of the New Mexico strain, no mice showed virus detection in the brain by 4 DPI; only half of the mice showed detectable virus in the brain at 7 DPI, while in another experiment, after challenge with 10[SUP]3[/SUP] PFU, virus was first detected in the brains of mice only at 6 DPI, in 4 out of 5 mice, with a mean titer of around 10[SUP]4[/SUP] PFU/g. Taken together, these results indicate that there is substantial variation in the pathogenicity among different isolates of the HPAI A(H5N1) currently circulating in dairy cattle in the United States. Further efforts should focus on determining the molecular determinants for the differences observed in pathogenicity and neurotropism among the different cattle H5N1 isolates, and the possible mechanisms for such.
One of the distinguishing characteristics of the current HPAI outbreak in dairy cattle is the fact that the mammary gland is the primary organ affected, leading to the production of infected milk containing high titers of infectious viruses, which may pose a severe risk to people through the consumption of raw milk, which is emphasized by the occurrence of dead peridomestic animals on dairy farms affected, attributable to the consumption of raw milk. With that in mind, we explored the pathogenicity o cattle-HPAI in mice through the oral route. Eisfeld et al. (2024) reported a similar experiment; however, they administered the inocula by pipetting to the back of the throat. Considering that the oropharynx is also part of the upper respiratory system, direct respiratory infection by such methodology cannot be ruled out, a fact further stressed by the report that, after inoculation, mice showed milk in the nasal cavity[SUP]34[/SUP]. Therefore, we chose to inoculate the mice through oral gavage, to ensure challenge through the digestive system while preventing the possibility of direct respiratory infection. Cattle-HPAI retained its high pathogenicity also after oral challenge, which is consistent with previous reports[SUP]17,18[/SUP]. Although we did not perform challenge of mice with M244/05 through oral gavage, it has been reported that that virus is also capable of infecting mice through that route[SUP]14[/SUP].
We also directed efforts towards the characterization of the interactions of the cattle H5N1 virus with mammalian host cells in vitro. We found that cattle-HPAI has extremely high virus fitness in at least three human lung cell lines —A594, Calu-3 and MRC-5—growing to two log[SUB]10[/SUB] higher titers at 48 HPI than M244/05. This is consistent with a report from Gu et al., 2024, who found that a cattle-derived H5N1 isolated from a human patient in Texas grew faster than A/Vietnam/1203/2004 in human primary lung alveolar cells, reaching very high titers. This finding may have important implications for human health: even though all reported cases of H5N1 infections in humans acquired from exposure to cattle have been mild, severe outcomes in case of lung infection by the virus might seem likely in the unpredictable future. This is further emphasized by recent cases of severe human H5N1 infection in Canada and the United States by viruses of the clade 2.3.4.4b, associated with the ongoing poultry outbreak of HPAI[SUP]35,36[/SUP].
Using human lung Calu-3 cells overexpressing either RIG-I or MDA5, we also found that the growth capacity of cattle H5N1 virus is unabated, in contrast to that of M244/05 virus, suggesting a more efficient interferon antagonism may account at least in part for its superior growth kinetics and, possibly, pathogenicity. Interestingly, despite the resistance to interferon signaling observed in vitro, we did see an induction of type I interferons at the transcription level in both lungs and brains of mice infected with cattle-HPAI, in keeping with the upregulation of the pro-inflammatory cytokines tested, as is typical of a cytokine storm associated with highly pathogenic virus infection. A similar scenario has been observed with other highly pathogenic viruses – for instance, while Ebola is known to antagonize the interferon system by a variety of mechanisms[SUP]37,38,39,40[/SUP], highly elevated expression of interferons is nevertheless a hallmark of Ebola infection[SUP]41,42,43[/SUP]. It is conceivable that an inhibition of antiviral signaling may confer an advantage to virus replication at an early stage leading to overwhelming infection, before an exuberant inflammatory response at cytokine level takes over, leading to the indiscriminate overstimulation of inflammatory pathways (cytokine storm scenario). Further efforts towards the elucidation of the molecular mechanisms involved in this innate immune antagonism are warranted.
...
https://www.nature.com/articles/s41598-025-93493-5